136 University of Ransas Geological Survey. 



since there is no sacrum in these animals. I know of no reptile, 

 as I have previously stated, in which this arrangement of the 

 non-costiferous vertebras is not the rule, and it seems strange 

 that the error of confounding the non-costiferous and non-chev- 

 ron-bearing vertebras with the mammalian lumbar vertebrae 

 should have persisted so long. In the Mosasauria there is never 

 a sacrum ; the rod-like ilia are directed, not backward, but for- 

 ward, ending without attachment, but doubtless lying in conti- 

 guity with the most anterior of the pygals. The symphysis of 

 the ischia is thus thrown below the fourth or fifth vertebra suc- 

 ceeding. If these vertebras bore chevrons it will be immediately 

 seen they would have protruded into the pelvic cavity, obstruct- 

 ing the outlet. Not less than five or six pygial vertebras are 

 necessary in these marine lizards to leave space for the free exit 

 of the cloaca. 



The relative lengths of the thorax and tail vary in the different 

 genera. While in Clidastes velox there are as many as forty- 

 two precaudal vertebras, in Tylosaurus and Platecarpus there are 

 not more than thirty, with seventy-seven caudal vertebras in 

 the first and eighty-six or more in the last two. 



Atlas. 



Clidastes velox. The intercentrum is a small bone with its 

 three principal sides of nearly equal extent ; the two upper 

 articular surfaces are gently concave, and meet in a transverse 

 obtuse border. The anterior margin is thin and sharp, the 

 articular surface behind it for the condyle more concave than 

 that for the odontoid. The inferior surface is convex, both 

 antero-posteriorly and from side to side, with a roughened 

 longitudinal prominence in the middle on the posterior part for 

 muscular insertion. The posterior articular surface is limited 

 by a narrow groove below on the margin of the bone. The 

 bone articulates broadly behind with the atlantar hypapophysis, 

 reaching above to the surface for the odontoid, and, on the ends, 

 to those for the lateral pieces. The lateral pieces have dis- 

 tinctly separated articular surfaces for union with the condyle, 

 intercentrum, odontoid, and axis. The facet for the intercentrum 



